There is a pressing need for new approaches in the treatment of glioblastoma multiforme (GBM). These new approaches will likely encompass both new therapeutic agents, and novel drug delivery regimens that seek to reduce toxicity and change the focus from tumor eradication to tumor management. We have shown that the synthetic thymidine analog, 5-ethynyl-2'-deoxyuridine (EdU), has profound anti-cancer activities (Ross, et al., 2011), and preliminary evidence suggests that EdU in an adaptive dosing paradigm can establish long-term stasis of xenografted human GBM tumors. Standard chemotherapy attempts to eradicate all cancer cells with drug doses that are near the maximum tolerated by the patient. This approach kills many cancer cells and can effectively prolong survival, but suffers from two serious problems: first, it nearly always selects for drug-resistant cells, allowing them to dominate the tumor; second, it contributes to poor quality of life by making patients quite ill. The theory behind adaptive therapy (borrowed from ecological population dynamics; Gatenby, et al., 2009; Gatenby, 2009; Wolkenhauer, et al., 2010) is that most continuous, high-dose drug regimens will eventually lead to the appearance of resistant cells that dominate the tumor. However, resistant cells are less fit (i.e. divide slower) than no-resistant cells in the absence of drug treatment; that is, in order to become resistant a cell must pay a cost by sacrificing other functions, and this sacrifice is manifested as a slower expansion. In the absence of drug, the non-resistant cells will outcompete and suppress the resistant cells by expanding faster and dominating the resources of the tumor. However, in the presence of high doses of drug the reverse is true. In this way tumors eventually consist only of resistant cells, and become refractory to treatment. Adaptive therapy adjusts drug dose according to the tumor response; if the tumor grows, then the dose is raised, but is lowered again when the tumor stabilizes or shrinks. The hope is to kill enough non-resistant cells to prevent the tumor from growing uncontrollably, but not so many that the constraints on the non-resistant cells are removed. In this way, the tumor will always consist of drug-sensitive cells that can -with dynamic dosage adjustments - be trimmed and managed over time, though it will likely not be completely eradicated. Here we propose to: 1) optimize an adaptive EdU protocol for the treatment of human GBM using mouse xenografts; 2) compare the efficacy of the adaptive EdU protocol to a standard therapy approach in the treatment of human GBM in mouse xenografts; and 3) assess potential negative consequences of EdU administration on the neural and hematopoietic stem cell niches. The overarching goal of these studies is improve treatment options for human glioma by testing the potential of an adaptive therapy approach using a novel chemotherapeutic candidate. In addition to introducing a new anti-cancer compound, our proposal has the potential to radically change dosing paradigms for a variety of existing treatment options.